Our superannuation system

I compiled a list of thoughts about our own superannuation system in response to a journalist from elsewhere thinking about pensions in their own country and asking me for a rundown on Australia’s system.

Via various accords with their union and business partners, the Hawke/Keating Government brought all regular employees within the super net, requiring compulsory contributions of 10% of payroll — up from 9.5% on 1 July this year. It will rise by 0.5% to a target of 12% over the next four years.

• defined benefit schemes became of less and less importance and the whole compulsory system is built around defined contribution — I think the Commonwealth Government has also been winding down its defined contribution scheme towards something that looks more like a defined contribution scheme, though I expect it contains hybrid elements.
• While some tax and regulatory incentives encourage people to take annuities, this is nevertheless funded from defined contributions. Beyond that, I don’t think it delivers any ‘defined benefit’ attributes of the old schemes.

Here are some things I think there are to like about the scheme:

• It’s provided a good way of getting savings heavily into equities. A lot of our funds spend a lot of their time about 75% long on equities, including substantial investment in foreign equities which is good from a risk perspective.
• It provides a way for a substantial amount of self-provision and self-management of investment — which is fine if it’s kept reasonably ‘vanilla’ flavoured — of which more below
• It provides a mechanism for cranking up national savings — which Australia was very worried about in the 1980s — and it has broad political support — not something that a lot of savings initiatives have!
• I argued here that the size of our indigenous savings pool plays some role in Australia’s substantial out-performance of New Zealand since the 1980s but the evidence for this isn’t overwhelming.

Things not to like:

• There’s a complicated set of tax concessions that seem to me to be unnecessary and inequitable when one is running a compulsory scheme. Compulsion is the ultimate incentive!  In any event, most taxes on the scheme are flat at 15% or zero during the drawdown phase — which of course creates a regressive carve out of the otherwise progressive income tax scheme for a vast pool of savings (currently over $2 trillion and growing quite fast still). I’m told by a super expert that the largest family super fund has more than $500 million in it — all taxed concessionally for a single family! The Turnbull Government increased taxation on the wealthiest superannuants.
• Costs of funds management in the system are several times what they should be, though they’ve been coming down owing to policy change
• The regulatory regime of the family super schemes is both ridiculously complex and cruelly dysfunctional.
• The switch has been thrown too far towards defined contribution leaving little risk pooling in the system. I think we should think of taking a leaf out of Shiller’s book and creating some government instruments designed to allow defined contribution schemes to insure against a range of risks that the market either doesn’t provide or provides inefficiently. These include
  • Intergenerational risk — a product of the kind I sketched in this column.
  • Longevity risk
  • End of life health risk.
• These latter two risks can be insured against now but the private market will struggle to provide them efficiently. Mind you one can imagine ways in which markets could be helpful in encouraging healthy behaviour, though private insurance markets have a pretty poor record in doing so.
• Particularly because the super lobby is now so politically powerful, it has even been able to head off proposals for people to access some of their savings for a deposit on a home. It is obviously sensible to allow people to invest their savings in a home since returns on property can be expected to be relatively high, it diversifies the savings portfolio from equities and there are various additional benefits from home ownership over renting as well as external benefits to neighbourhoods.

One of the reasons I’ve always been in favour of removing all tax concessions in a compulsory system is that these concessions only make sense if they’re

• Well-crafted
• Stable over time
• Simple.

As I feared at the time and as has been born out in Australia, none of these conditions hold. Tax concessions tend to get changed quite often by governments, which face poor incentives to get it right beyond the optics of the existing parliamentary term. And even if they do get it right, there’s usually a lot of grandfathering for good and bad reasons as you go. In no time no normal person can make decisions in such a system — you need a professional to steer you through it all.

I’d also be happy to revise my view if I were shown contrary evidence, but I’m basically against savings incentives generally. Come prudential compulsion may make sense — but I’m pretty suspicious of special incentives to save for these reasons:

1. They’re regressive, going preponderantly to the wealthy;
2. I doubt the elasticities are high so they’ll come with a lot of deadweight loss; and
3. Short of a progressive consumption tax which I don’t think has been properly implemented anywhere, one can expect a substantially higher deadweight loss because special savings vehicles attract savings to a substantial extent by diverting them from elsewhere — I’m pretty confident all of that $500 million sitting in a self Managed Super Fund would have been saved without the incentives. I doubt it’s been scrimped and saved in response to the incentives.

Given all these complexities, I also think what I call ‘extended competitive neutrality’ can be very useful. Governments run super schemes for their employees, and given that they do that, there’s no reason not to open them up to anyone who wants one. This is a very simple hack, but that somewhat undersells its attractions. Again and again, policymakers — here and I think elsewhere — think the answer to market failure is to try to make the area of policy look more like a functioning market. So they leave people ‘free to choose’ and then regulate for better information flow.

The problem with that is not just that regulating for good information flow is much easier said than done (and in many ways, we don’t really try — paying vanishingly little attention to the main vector by which people inform themselves in complex markets which is not by doing ‘due diligence’ themselves but by seeking out a competent professional advisor — if we took that problem seriously regulation would look quite different). But even more fundamentally, there are always large shares of the population (rarely less than 40% but often up to 80% or more) who just don’t feel competent to manage an investment portfolio. While I think those who are comfortable should be able to have that choice, it seems to me not just highly inefficient but actually cruel to hand such people over to a market they neither understand nor want to understand.

We could deal with a great deal of the trouble experienced by ordinary folk by ensuring that, wherever it’s easy we should give ordinary people a ‘public option’ of a properly run, appropriately priced version of any expert service they need. We do it in health and education. Nor does this have to be by public provision. It can be by publicly supervised tendering of services.